The adult respiratory distress syndrome (ARDS) is one form of acute acute hypoxemic respiratory failure (AHRF) in which acute lung injury causes pulmonary vessels to leak plasma into alveolar spaces causing intrapulmonary shunt (Qs/Qt) and arterial hypoxemia which is refractory to O2 therapy. Since no effective prevention or specific therapy of the pulmonary capillary leak is yet available, this proposal seeks to improve the principles of supportive therapy by defining the mechanisms of three interrelated and poorly understood problems in canine models of AHRF: 1) Relationships among pulmonary vascular pressures, flow, shunt, and edema. Extracorporeal membrane oxygenation (ECMO) will be used to reduce pulmonary pressure and flow after acute lung injury induced by intravenous oleic acid and by intrabronchial HCl. A similar experimental approach will determine the effect of increased hematocrit on pulmonary edema accumulation. If ECMO and increased hematocrit do not increase edema but do improve peripheral O2 delivery (QO2), both represent new therapeutic interventions in AHRF. To determine the mechanism whereby acute lung injury reduces pulmonary blood flow in edematous areas, ECMO will also be utilized to provide partial bypass of the lung to define the pulmonary vascular pressure flow relationships before and after acute lung injury. 2) To explore mechanisms accounting for pathologic supply dependency of oxygen consumption induced in dogs by high F1O2, we will extend our prior studies of the limits of aerobic metabolism by studying regional vascular beds, specifically the gut and hind limb during ventilation with room air or O2. These studies should elucidate whether and how hyperoxia redistributes limited QO2 among peripheral parallel organ systems in an inefficient manner. The same methodology will be used to compare the limits of aerobic metabolism when QO2 is reduced by progressive hypoxia, by anemia, or by progressive reductions in blood flow. 3) To test the possibility that ventricular pumping function is depressed in AHRF due to hypoxia, acidemia, or both, we will measure left ventricular pressure dimension relationships, coronary blood flow, and myocardial oxygen extraction during progressive hypoxia. We will use this methodology to determine the effects of vasoactive drugs on bacteremia, hyperoxia, and acidosis on these limits of myocardial aerobic metabolism and to determine how a reduced resistive afterload improves depressed myocardial pumping function in pentobarbitol anesthetized dogs. The results of these canine experiments should elucidate important mechanisms influencing the pathophysiology and treatment of pulmonary capillary leak and its systemic consequences on the gut, hind limb and heart, and facilitate clinical trials to improve supportive therapy of AHRF.